ReSpinQuant: Efficient Layer-Wise LLM Quantization via Subspace Residual Rotation Approximation 文章

ArXiv CS.CV2026-05-29NEWSen作者: Suyoung Kim, Sunghyun Wee, Hyeonjin Kim, Kyomin Hwang, Hyunho Lee, Nojun Kwak

ReSpinQuant: Efficient Layer-Wise LLM Quantization via Subspace Residual Rotation Approximation · 相关事件